Abstract
Six techniques of proximal femoral replacement were compared in vitro with the use of compression, bending, and torsional testing in a canine model. One femur of each pair was osteotomized in the midshaft region, and the proximal portion was replaced with one of six techniques. These techniques included (a) a segmental proximal femoral endoprosthesis cemented into the distal femur with no allograft (technique E~s~); (b) a long‐stem endoprosthesis press‐fit into an allograft and cemented into the distal femur with a transverse osteotomy (technique A~T~); (c) the same construct as technique A~T~, but with a step‐cut at the osteotomy (technique A~S~); (d) a long‐stem endoprosthesis interlocked into an allograft and cemented into the distal femur with a transverse osteotomy (technique A~I~); (e) a short‐stem endoprosthesis cemented into an allograft combined with one plate laterally stabilizing the allograft to the distal femur with a transverse osteotomy (technique A~P1~); and (f) the same construct as technique A~P1~, but with an additional plate cranially (technique A~P2~). A long‐stem endoprosthesis cemented into the contralateral intact femur served as the control. Techniques that involved a long‐stem endoprosthesis and cementing distally (A~T~, A~S~, and A~I~) were more resistant in torsion than the plated replacement techniques (A~P1~ and A~P2~). The segmental replacement construct (E~S~) was equal to or stronger than all other techniques under each testing condition. In torsion, the addition of a step‐cut (A~S~) significantly lowered angular displacement of the reconstruction when compared with the reconstruction with a transverse osteotomy (A~T~) (p ≤ 0.05). Bones with one‐plate fixation (A~P1~) were significantly weaker in torsional stiffness and maximum torque and in mediolateral bending (p ≤ 0.05) than all other techniques. The addition of a second plate (A~P2~) increased the mechanical properties of the construct so that it was greater than the one‐plate method and was equal to (bending and compression) or still weaker (torsion) than the other techniques. The results indicate that segmental replacement methods and allograft/endoprosthetic composites that involve long‐stem endoprostheses fixed with cement are mechanically superior to methods that involve short‐stem endoprostheses with single or double plating at the osteotomy sites.